Arsenic trioxide (As2O3) induced complete remission in acute promyelocytic leukemia (APL, AML-M3) patients that relapsed after all trans retinoic acid (tRA) and chemotherapy treatment. Clinical results indicated that the therapeutic effect of As2O3 in APL correlated with the expression of PML-RARalpha , the product of the t(15;17) translocation, and was mediated by apoptosis and non-terminal differentiation induction. We have found that As2O3 degraded PML-RARalpha and allowed RARalpha (from the wild-type allele) to drive APL cell partly differentiation. However, the connection between PML-RARalpha expression on one hand, and apoptosis induction by As2O3 on the other hand, is unclear. We have found that 1) APL cells contained low amounts of glutathione-s-transferase pi (GSTpi), glutathione peroxidase (GPx), catalase and high amounts of myeloperoxidase (MPO); 2) APL cells were highly sensitive to As2O3-induced apoptosis in vitro by a hydrogen peroxide (H2O2) mediated pathway; 3) Ascorbic acid selectively increased As2O3-induced apoptosis in HL-60 cells (which express high amounts of MPO) not in U937 and normal bone progenitors cells (which do not express MPO). We hypothesize that 1) low levels of GSTpi allow As2O3 to inhibit GPx. GPx inhibition in combination with low catalase expression will result in H2O2 accumulation; 2) accumulated H2O2 is converted into reactive oxygen species by MPO, and then trigger apoptosis; 3) PML-RARalpha sensitizes APL cells to As2O3-induced apoptosis by upregulating MPO and/or downregulating GSTpi, catalase and GPx; 4) Ascorbic acid selectively synergizes As2O3-induced apoptosis in MPO positive AML cells by producing H2O2 and depleting reduced form glutathione (GSH), the substrate of both GSTpi and GPx. The initial aim of the project is to confirm that As2O3 induces apoptosis through H2O2-mediated pathway. This will be tested by comparing H2O2 amount and apoptosis induction in As2O3 treated AML cells. The second aim will determine the central role of GSTpi to control the sensitivity of cells to As2O3-induced H2O2 accumulation and the third aim will examine the functions of MPO in sensitizing As2O3-induced apoptosis. These will be tested by stably transfecting sense or antisense cDNA and using specific inhibitors. The fourth aim will dissect the connection between PML-RARalpha expression and the levels of GSTpi, GPx, catalase and MPO. PML-RARalpha stably transfected cells will be used for this purpose. Our last aim will evaluate the selective apoptosis-induction and the mechanism of As2O3 in combination with ascorbic acid among AML cells with/without expressing MPO in vitro. SCID models bearing AML cells will be used to test the in vivo effect. Successful completion of the proposed studies will not only contribute to elucidation of the mechanism of As2O3-induced remission in APL, but may also provide innovative usage of As2O3 in other forms of AML.